Observation Modes

Important information for CfP S27A

We are calling for community fillers in S27A.
Telescope downtime The telescope will be unavailable for maintenance during two periods: (1) from February 1 to mid-March for AZ/EL incremental encoder replacement, and (2) beginning in mid-July for primary mirror recoating. Please make observation plans with careful checking of target visibility.
The priority of flux standard stars
Guaranteed time observations (GTO) will be allocated several nights in S27A.

For investigators in S27A

Queue mode

Queue mode is the default observation mode for PFS.

In Queue mode, the investigator will not attend the observations. They will be carried out by the Subaru staff. The observation plan (including telescope pointing and fiber allocations) will be optimized by proposal rank, target priority, and target visibility to maximize the observation efficiency. The highly-ranked proposals will have higher priority in determining the telescope pointing and the fiber position, while the unused fibers can be assigned to other, lower-ranked, programs. Note that only highly-ranked proposals can, in principle, drive the telescope pointing.

According to the decision of Subaru Advisory Committee (SAC), 30% bad weather factor will be considered for time allocation of the Queue mode programs in order to increase the completion rate by ensuring sufficient observation time.

Important Queue mode policies for PFS

The fibers will be primarily allocated to accepted Queue mode programs. After that, any remaining fibers will be assigned to filler targets, which may include non-accepted Queue programs, community fillers, and observatory fillers.
Target and observation configurations cannot be adjusted after the proposal submission.
The exposure time for a single frame is fixed to 7.5 min (= 450 sec), and two exposures with the exposure time of 7.5 min will be taken for each pointing with the same fiber design, i.e., total exposure time for each fiber design is 15 min (= 450 sec x 2 frames). This is a technical requirement for the cosmic ray rejection. See also here for details.
All Queue mode targets must be fainter than 16 mag (AB mag) over the entire PFS wavelength coverage to avoid detector saturation and to reduce near infrared persistency effects.
The investigators request their observing time for PFS Queue observation in units of "Requested Observing Time (ROT)". ROT is the time needed to complete the program by itself without sharing fibers with others. The ROT is calculated as "Number of PFS Pointing Centers (PPCs)" × (15 min + overhead).
- ROT does not depend on how many fibers are assigned in a single PPC.
The maximum ROT for a normal program is 35 hours (equivalent to 3.5 nights) within a semester. The ROT for an intensive program is limited to 280 hours (28 nights) over six consecutive semesters and to 140 hours (14 nights) per semester.
Along with the ROT, the "Fiber Hour (FH)" is defined, which is the sum of the exposure time needed for each target.
- ROT and FH will be calculated when uploading the target list to PFS Target Uploader.
The observation time is allocated in FH. The allocated FH (FH_alloc) can be reduced by the observatory, depending on the score of the proposal.
The effectiveness of the exposure is evaluated in terms of "Effective Exposure Time (EET)". For details on EET, see also here. EET is the S/N-based estimation evaluated with respect to the nominal condition.
- For example, if a 15 min exposure is taken with relative throughput=0.8 and relative noise=1.2 with respect to the nominal condition, the effective exposure time is calculated as (0.8 / 1.2)² × 15 = 6.7 min.
- If the observing conditions are better than nominal, a 15 min exposure may be counted as, e.g., 20 min. If the conditions are poor instead, it can be, e.g., 5 min.
- However, when the sum of the EET for a target exceeds the required exposure time for that target, then the observatory does not charge the excess time to the program. For example, if the EET becomes 105 min for targets requesting 100 min, the extra 5 min is not charged to the program.
The completion rate of the program is calculated by Σ_i EET_i / FH_alloc, where Σ_i EET_i is the sum of the EET of all targets. Completion is defined as 100% when the observed FH matches the allocated FH. The cumulative EET of individual targets are not accounted for.
The Queue proposal will be classified into the following categories.
- Accepted
- Non-accepted (used as a filler with higher priorities compared to community/observatory filler)
- Rejected (never observed)
No carry-over, regardless of the category of programs.

Classical mode

In Classical mode observations, at least one investigator must attend the observations. Investigators may apply for Classical mode only if their requested observations cannot be accommodated within the current Queue mode scheme.

Please check Justifying Classical Mode Observations, Comparison Classical vs. Queue and Acceptable cases carefully.

See also Proposal preparations for Classical mode.

Justifying Classical Mode Observations

Justifying Classical Mode Observations.

Proposers must clearly explain the reason for requesting Classical mode observations to achieve their scientific goals in the proposal. During the review, the Time Allocation Committee and the observatory will judge the necessity of Classical mode. If they consider that the program can also be achieved in Queue mode, it can be assigned as a Queue program.

Comparison between Classical and Queue modes

Category	Classical Mode	Queue mode
Allocation & Recovery	Observation time is allocated as specific nights and hours. There is no compensation for downtime caused by bad weather or unforeseen incidents.	All accepted programs include a 30% weather factor in the accepted ROT. The observatory aims to ensure observed FH matches 100% of allocated FH.
Flexibility	Highly competitive with a strictly binary outcome ("accepted" or "rejected"). No possibility of being saved as a filler.	High probability of re-planning for the same targets depending on priority or visibility, even if unforeseen incidents or bad weather occur.
Review & Acceptance	Proposals are strictly reviewed regarding their "acceptable case" (justification for why Classical mode is required).	If a program is not accepted as a primary program, it may still be observed as a "filler" in the combination with other queue programs.
Post-Submission Changes	No additional targets allowed. However, adjusting priority, exposure time, or deleting targets is acceptable until ~4 weeks before the run.	No adjustments to target information or observation configurations are permitted after the proposal submission.
Preparation Phase	Investigators can iterate fiber designs a few times with the observatory during the preparation phase.	No iteration. The observatory manages the queue execution based on the submitted configurations.

Acceptable cases of Classical mode

Acceptable cases of Classical mode (last update: Dec. 2025)

The target list can only be fixed after the proposal deadline (e.g., Target of Opportunity (ToO))
The target list can only be fixed after the proposal deadline but observation field can be fixed (e.g., ToO in already known fields)
Time-critical observations
A program that needs to repeat several exposures with the exact same fiber allocations for long hours (e.g., > 1 hour)
Individual total exposures have to be different from those of queue mode (450 sec x 2 frames) in the case where clear scientific advantage is recognized
The pointing centers have to be fixed by the investigator(s) in the case where clear scientific advantage is recognized
A program requiring to use the IFU-like mode (since S26A) see for details
A program requiring a special calibration target list to observe at crowded fields, like the galactic center, clusters, or nearby galaxies (e.g., M31).
Even if the target field lacks observatory calibrators, the program could potentially be moved to Queue (by uploading calibrators).

Even with Classical mode, the complete target list must be submitted via PFS Target Uploader before the proposal submission, except for case 1 shown above. It is not allowed to update the target list during the observation preparation phase.

During the Classical mode observations, investigators can set/change the exposure time and repeat the exposure with the same fiber allocations, while the observatory determines the fiber allocations. The pointing centers of the telescope will also be determined by the observatory except for case 6 shown above (the observatory still sets the fiber allocations). This is to efficiently use the open fibers for other targets such as community fillers and observatory fillers.

Target of Opportunity (ToO)

ToO observations with PFS will be conducted as Classical mode observations. Therefore, the observers must attend the observations. PFS is accepting three types of ToOs: non-rapid ToO, rapid ToO, and transient observations. The ToO trigger must be accepted by the observatory beforehand by following the same procedures as for other instruments (see here), and the ToO targets have to be uploaded via PFS Target Uploader before triggering the ToO. For non-rapid ToO, the trigger deadline is 9 A.M. HST one day before the observing night, and for rapid ToO the deadline is 9 A.M. HST on the day of the observing night. In case the PI wants to propose a rapid ToO, they MUST contact the observatory before the proposal submission. Open fibers may be used for Queue program targets, community fillers, and observatory fillers.

PFS ToO observations can only be triggered on the nights when the PFS observations are scheduled. In the semester S26B, the execution time of ToO observations will be in a unit of 0.5 nights.

Please review the current ToO and Transient policy before submitting a ToO proposal.

See also Proposal preparations for ToO.

IFU-like mode

The IFU-like mode is a PFS observational method available only in Classical mode. It enables two-dimensional spectroscopic data acquisition for every science target, using multiple dithered fiber positions to simulate the functionality of an Integral Field Unit (IFU).

Only simple observation setups will be accepted. For instance, we currently anticipate limiting the available fiber allocations to two predefined patterns, such as triangular and hexagonal arrangements. More flexible configurations may be considered in future semesters.

See also Proposal preparations for IFU-like mode.

Specifications of IFU-like mode for S26B:

Investigators should always contact pfs-obs-helpdesk before the submission deadline to check the feasibility.
IFU-like mode enables sequential observations (IFU sequence) of both the center and surrounding areas of each target, similar to standard dithering techniques. We call the fiber positions surrounding the center target position the variant fiber position.
The maximum exposure time at variant positions is 30 min (900s x 2) in a single IFU sequence because a fiber reconfiguration at least every 30 min is required to ensure the accurate fiber positions.
If a PI requests over 30 min exposures (in total) at each variant position, multiple sequences are necessary. It takes the same time as the previous IFU sequence.
In the IFU-like mode, variant fiber positions can be specified only in a regular triangular or hexagonal pattern with a rotation in the range of +/- 30 deg (for hexagon) and +/- 60 deg (for triangle) at a fixed separation around each target's central coordinates.
Separation of variant fiber positions from the center position can be specified (tested up to 1.0").
Data can be processed by the PFS DRP system, and PI will receive the resultant 1D spectra in the Science Platform, the same as other observation modes.
The positional accuracy of the variant fibers is comparable to the standard fiber alignment.

Fillers

The filler targets for PFS are used to fill up the open, unassigned fibers in the Queue/Classical/ToO observations.

Important: PFS fillers will not affect the telescope pointing!

PFS filler is a completely different category from HSC fillers, which are used to fill the observation time in poor weather conditions. PFS fillers will be used only to assign targets to surplus fibers, and therefore the telescope pointing will not be determined by the fillers. Basically, a filler target that is located at a field where no other Queue/Classical/ToO programs look at will not get observed.

Community fillers

We ARE CALLING FOR proposals of community fillers for S27A semester.

Investigators can submit a proposal for community fillers to request observations as a filler. An international call for community filler proposals is issued every two semesters, and the accepted programs run for two consecutive semesters.

The duplicated submission of the same proposal to the normal program and the community filler program is not allowed.
The investigator must explain the reason for applying for the community filler and why it cannot be done with the normal program.
The proposals are reviewed by the distributed peer review process, i.e., investigators (either the PI or the Co-I) review each other’s proposals.
An upper limit on the target sample size and source brightness may be imposed for technical reasons.
- If the investigator requests more than 100K targets, the observatory may ask the investigator to reduce targets for technical reasons. Prepare the proposal with this possibility in mind.
- All targets have to be fainter than 17th magnitude over the entire PFS wavelength coverage. We will make the magnitudes cut for community fillers before observation planning.

The target of the accepted program has a standard proprietary period (1.5 years). Investigators cannot access the raw data during the proprietary period, but will receive the fully reduced and calibrated spectra data through the PFS science platform.

If a target included in a community filler program is also targeted in accepted normal programs or other community filler programs that have higher priorities, it will be excluded from its list. The investigators of this filler program cannot access to the data even if the target is observed.

See also Proposal preparations for community fillers.

Observatory fillers

Observatory fillers are targets that are prepared by the observatory, to fill the remaining fibers after assigning community filler targets. The PS1 sources with i_P1 = 18–21 mag will be used in Observatory filler. If a source is targeted in accepted normal proposals or community filler programs, it will be excluded from the observatory filler program.

The fully reduced and calibrated spectra of the observatory filler targets will be made public after every observing run or every semester from the PFS science platform. The raw data will be released to the public after 1.5 years.